Adaptive Medical Image Presentation for Anomaly Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current automated medical imaging analysis systems face challenges in efficiently detecting anomalous objects, such as solitary pulmonary nodules, due to high false positive and false negative rates, requiring extensive manual inspection by radiologists, which is time-consuming and prone to subjective interpretation.
Innovation Solution
A system and method for automatically generating an adapted presentation of candidate anomalous objects in medical images using a detection classifier and a presentation parameter classifier, optimizing the presentation for visual inspection by computing parameters based on the location and neighboring anatomical features to improve detection accuracy and reduce false positives/negatives.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If automated analysis systems are used to detect anomalies in medical images, then detection speed is improved, but false positive and false negative rates increase
Solution Approach 1:
The patent introduces presentation parameters as an intermediary between the automated detection system and the radiologist. These parameters optimize the visual presentation of candidate anomalies, serving as a mediator that bridges the gap between automated detection and human verification, thereby maintaining high detection speed while improving reliability through better visualized evidence.
Solution Approach 2:
The system dynamically changes presentation parameters (such as window width, window level, image contrast) based on the detected candidate anomaly characteristics. This parameter optimization allows the system to maintain high detection speed while improving the reliability of anomaly identification by presenting visually optimized images that reduce false positives and negatives.
2Reliability
If manual inspection by radiologists is performed to verify anomalies, then detection accuracy is improved, but time consumption increases
Solution Approach 1:
The system performs preliminary optimization of image presentation parameters before the radiologist conducts manual inspection. By pre-optimizing the visual presentation based on detected candidate anomalies, the system reduces the time radiologists need to spend on verification while maintaining high detection accuracy, as the images are already prepared in the optimal viewing configuration.
Solution Approach 2:
The system extracts and highlights only the relevant candidate anomaly regions with optimized presentation parameters, removing unnecessary visual information. This allows radiologists to focus their manual inspection on the most suspicious areas with optimized visualization, reducing overall inspection time while maintaining high detection accuracy.
3Reliability
If extensive manual inspection is performed to reduce false positives, then detection reliability is improved, but productivity decreases
Solution Approach 1:
The system automatically adjusts presentation parameters (contrast, brightness, window settings) based on the characteristics of detected candidate anomalies. This parameter optimization enhances the visual distinctiveness of true anomalies while suppressing false positives, allowing radiologists to maintain high throughput while improving reliability through better visual differentiation.
Solution Approach 2:
The patent replaces the mechanical process of extensive manual inspection with an automated parameter optimization system. This substitution maintains high detection reliability by systematically optimizing visual presentation to reduce false positives, while preserving productivity by eliminating the need for time-consuming manual re-adjustment of viewing parameters.
Data Source
AI summary
There is provided a computed implemented method of automatically generating an adapted presentation of at least one candidate anomalous object detected from anatomical imaging data of a target individual, comprising: providing anatomical imaging data of the target individual acquired by an anatomical imaging device, analyzing the anatomical imaging data by a detection classifier for detecting at least one candidate anomalous object of the anatomical imaging data and computed associated location thereof, computing, by a presentation parameter classifier, at least one presentation parameter for adapting a presentation of a sub-set of the anatomical imaging data including the at least one candidate anomalous object according to at least the location of the candidate anomalous object, and generating according to the at least one presentation parameter, an adapted presentation of the sub-set of the anatomical imaging data including the at least one candidate anomalous object.